Mechanism by which positive end-expiratory pressure increases cerebrospinal fluid pressure in dogs.

We investigated possible mechanisms by which positive end-expiratory pressure (PEEP) increased cerebrospinal fluid pressure (PCSF) in anesthetized mechanically ventilated dogs. In part I of the study, PEEP was applied in 5 cmH2O increments each lasting 1-2 min, before and after a snare separated the spinal from the cerebral subarachnoid space in each animal. Next, with the spinal cord still ligated, the dogs were ventilated without PEEP while superior vena cava pressure (PSVC) was raised in 5 cmH2O increments by means of a fluid reservoir connected with the superior vena cava. Cerebrospinal fluid pressure in the cisterna magna increased immediately and in parallel with PEEP before and after the spinal subarachnoid space was occluded and also increased when PSVC was raised independently; in all circumstances the increase in PCSF correlated closely with PSVC (r = 0.926). In part II of the study, arterial blood gases were drawn before and after PEEP was applied in the same increments and for the same duration as in part I. Cerebrospinal fluid pressure measured with a hollow skull screw again rose in parallel with PEEP, whereas arterial carbon dioxide tension rose only slightly at 60 s. In part III of the study, mean arterial pressure (Pa) was allowed to decrease with PEEP or was held constant by distal aortic obstruction and volume infusion. Cerebrospinal fluid pressure increased regardless of Pa, but the increase was greater when Pa was held constant than when it fell with PEEP. We conclude that PEEP increases PCSF primarily by increasing PSVC and decreasing cerebral venous outflow. This effect is augmented if cerebral arterial inflow is increased as well.